Abstract:In group testing, simple binary-output tests are designed to identify a small number $t$ of defective items that are present in a large population of $N$ items. Each test takes as input a group of items and produces a binary output indicating whether the group is free of the defective items or contains one or more of them. In this paper we study a relaxation of the combinatorial group testing problem. A matrix is called $(t,\epsilon)$-disjunct if it gives rise to a nonadaptive group testing scheme with the property of identifying a uniformly random $t$-set of defective subjects out of a population of size $N$ with false positive probability of an item at most $\epsilon$. We establish a new connection between $(t,\epsilon)$-disjunct matrices and error correcting codes based on the dual distance of the codes and derive estimates of the parameters of codes that give rise to such schemes. Our methods rely on the moments of the distance distribution of codes and inequalities for moments of sums of independent random variables. We also provide a new connection between group testing schemes and combinatorial designs.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is in group testing, how to design a non - adaptive testing scheme to identify a small number of defective items among a large number of items while allowing a certain false - positive probability. Specifically, the paper focuses on constructing almost disjunct matrices, which can be used to design group - testing schemes so as to identify at most \(t\) defective items from a population of size \(N\), and the false - positive probability for each item does not exceed \(\epsilon\).
### Main contributions of the paper:
1. **Constructing almost disjunct matrices from non - binary codes**:
- The authors studied how to use non - binary codes with a large dual distance to construct almost disjunct matrices. By analyzing the distance distributions of these codes, the authors gave specific parameter estimations, such as specific applications for Reed - Solomon codes and algebraic - geometric codes (such as Hermitian codes and Suzuki codes).
2. **Constructing almost disjunct matrices from constant - weight codes**:
- The authors further studied how to use constant - weight codes with a known dual distance to construct almost disjunct matrices. By analyzing the moments of the distance distributions of these codes, the authors derived a specific relationship between the number of tests \(M\) and the false - positive probability \(\epsilon\).
### Specific results:
1. **Reed - Solomon codes**:
- For \(t> 18\log N\), the minimum is achieved by the first term; otherwise, it is achieved by the second term. For example, when \(t > 18\log N\), from equation (8) we can get:
\[
P_A(t, N)<\left(\frac{9e\ell^2t}{(q - t)^2}\right)^{\ell/2}\left(\frac{\ell}{2e}\right)^{\ell/2 + 1}\approx\frac{\ell}{2e}\left(2.13\ell^{3/2}\sqrt{\frac{t}{q - t}}\right)^\ell
\]
Therefore, if we take \(q>2.13\ell^{3/2}\sqrt{t}+t\), then the probability \(P\) is very small. For Reed - Solomon codes, we have \(M = q(q - 1)\), so the total number of tests \(M = O(\max\{t^2,t(\log_qN)^3\})\).
2. **Algebraic - geometric codes**:
- **Hermitian codes**:
- The parameters are chosen as \(n = q_0^3\), \(N\approx q_0^2q_0^2\), and the dual distance \(d'\geq q_0 + 2\). Assuming the number of defective items \(t < q_0^2\), then the number of tests \(M = q_0^5\).
- **Suzuki codes**:
- The parameters are chosen as \(n = q^2\), \(N = q^{2q_0^3+q_0 + 1}\), and the dual distance \(d'\geq2q_0 + 2\). Choosing \(r = 2q_0q\), then \(P\sim n^{-\ell/4}\) for any \(t\leq q/2\), allowing us to choose \(t = O(n^{1/2})=O(q)\). Therefore, the test matrix has an almost disjunct property, and the number of tests \(M = q^3\).
3. **Constant - weight codes**:
- The authors proved that if \(r < d'\), then the \(r\)-th central moment of the distance distribution is equal to the \(r\)-th distance distribution of a sphere of weight \(w\).